all right so welcome everyone to the very first process optimization academy session this is uh this is gonna be like one of those things you tell your friends you're at the very first one someday like being at shea stadium for the the first u.s based beatles concert or fenway park for carlton fisk's home run or some other super cool event that people recognize and relate to that i just didn't think of now the purpose of these sessions are to help us all get better educated on servicenow's in-platform process mining capabilities so we'll provide how-to's practical guidance and use cases product updates and as a group you know the idea is just simply to exchange ideas and collaborate so we all continue to get better so how will these sessions work we'll have a scheduled topic that we'll cover usually for the first 30 minutes today will be a bit longer probably since we'll be providing some foundational information as an intro to make sure that we level set everyone then we'll open it up for questions on the chosen topic for that session and then after that we'll open up for questions above and beyond today's topic so as i mentioned earlier these sessions will be recorded and eventually posted on the community so if that's a no-no for you what your your company you've been warned so consider yourself warned there and then finally uh feel free to post your questions and i said the queue i was going to use the q a window but i clearly set this up wrong so if you want to post your questions in the chat for today as we go through that's great often during the sessions i will pause and ask for questions when it makes sense other times i may ramble on for a bit and go back and check the the chat for questions at the end um you should all be on mute so if at some point we do pause and ask for questions just make sure you unmute yourself to ask that question and go back on you and i guarantee there will be times when we don't have an answer for your question immediately um but that's that's part of the purpose of these sessions is to flush out things that people are interested in understanding better than we just um weren't prepared to answer so in those situations what we'll do is we're going to follow up and and post them to the process optimization community that same place that you want to register or found the information to register for the session and then maybe what we'll do is we'll use you know for some of the more in-depth ones we'll use them as a topic for future sessions which is also something that we appreciate if you put in the chat this is the slide that says anything i say and do here today cannot be held against me in the court of law it's also the slide that says it as part of these conversations if we make any forward-looking statements about what may be coming in futurely releases you should take them as such make no decisions based on them because they're always subject to change and i'm assuming at this point you've read all the fine prints perfect so who am i uh my name is dan grady i'm an outbound product manager for our in-platform process mining solution or process optimization just a little bit about my background from a servicenow perspective i've been at servicenow for about six years now the the first uh four and a half of which i was focused on are now intelligent solutions so things like performance analytics predictive intelligence the virtual agent those were all things that i focused on helping customers better understand and get get the most value out of and then in the past year i spent some time on the creator workflows team which involves helping customers understand how they can use the platform capabilities to create applications above and beyond the out of the box applications that we offer using our low code tooling something we call app engine studio and then recently i joined this team of product managers focused on in-platform process mining and super excited about the opportunity because i think this really rounds out um the my skills on the platform as it comes as it pertains to let's say that hyper automation story and you'll see that as we go through the presentation the role that process mining plays in in hyper automation or a hyper automation plan if you will from an agenda perspective what we're going to be covering today and this looks like a lot but we'll move quickly as i mentioned earlier we need to set some foundational information out there as we move forward so we'll start off just with a little bit of the why and what of in platform process mining for those that are not familiar and then we'll talk a little bit about the process for optimizing a workflow it's it's much more than just process mining which is a critical component but there's there's an overall process if you think about continual improvement that you'll be going through and we'll do a quick demonstration of the analyst workbench which is what you see on the screen here so you get an understanding of kind of what the end user is going to experience whether it's the business process owner or the uh the business analyst that will be using the solution then we'll transition to getting an understanding of the building blocks that go into creating your model and we'll create your first process model and that's gonna by the end of this you'll have what you need to turn on um process optimization at your organization in the subprime instance and build your first model and get going and playing around if you like at that point we'll pause uh cover any questions that we hadn't got through the session we'll talk about some additional resources that are available to you above and beyond today and then we'll we'll touch on the future sessions that are coming this is something that we'll be running every other friday all right i always like to start sessions like this with a quote and i thought this one from the great american author maya angelou worked really well um do the best you can until you know better than when you know better do better it's why we're here today right to learn from each other and then go back and apply what we've learned and improve and this quote also aligns to the solution we'll be focusing on in these sessions moving forward right our process optimization solution which is the brand name we give our in-platform process mining capabilities that automate the process of x-raying our workflows and showing us where and how we can do better for anyone and everyone involved with them so we'll start with a little process mining 101 so with everything we do there's a design and desired path in our minds for how it should work and how it should play out whether that be planning an event like a wedding or these academy sessions which i will plan better for next time or a business process at an organization or even a park and when you design things you design for both efficiency as well as completeness to provide the best experience possible for everyone for example if you look at someone designing a park like what's on the screen here or like central park uh where i spent a lot of time i'm from new york they design it with paths for people to ride a bike on or walk a rollerblade on they put in a pond maybe so you can take a little bit a little row boat out and paddle around they put benches in for us to sit on read a book and share a story with a friend and there's a whole bunch of grassy areas for people to throw down a blanket have a picnic or play a game of scrabble on a saturday afternoon like my wife and i used to do right so we designed to provide the optimal experience for as many people as possible unfortunately what we design isn't always what ends up happening in reality and while this new path in reality may be more efficient for those trying to get from point a to point b in this scenario it certainly doesn't have a positive impact on the experience of my wife and i finding a nice cozy spot to play our scrabble game same thing plays out with our business processes designed for completeness and efficiency but the reality is that not all the work flowing through the process is taking that designed route which has a negative impact on the experience that both the end user as well as the people working the process are having there's a lot of inefficiency inside of it that we can remove and i defend identifying what is actually happening what's the truth in our business processes and then improving them isn't always as easy and obvious as that park example that we showed a little bit earlier for most of us continual improvement processes the continual improvement process is very manual time consuming expensive it involves pouring through tons of reports holding virtual and in-person workshops maybe hiring third-party consultants to come in and then ultimately getting alignment on the path to move forward right and at the speed that we're all moving right we're all moving at light speed these days even when we get through all of that and figure something out and have an idea in our head on how to move forward so much time has passed that the optimal path forward may have already changed the good news is with process optimization or in platform process money the opportunity is now there for us to do better in a few clicks we're going to be able to expose to relevant stakeholders the reality and truth of their workflows so they can start improving immediately and this is what you're going to learn today we're going to learn by the end of today the few clicks that you need to get started so process optimization can be applied to any process and provides visibility into bottlenecks it's going to shine a light on redundancies and rework it's going to point us towards specific automation opportunities that'll ultimately help us accelerate service delivery and optimize experiences it's going to help us increase automation rates it's going to help us make decisions around which automations or sometimes people and process adjustments will provide the that we that we can make that will provide the biggest impact and if you start taking a look at that list of questions the solution is going to help us answer things like where are we spending the most time which teams and groups are slowing us down what is the root cause behind some of the delays that we're experiencing these are questions we all have and we can now get to those answers in some cases in a matter of minutes i was mentioning earlier i've spent a number of years in the servicenow ecosystem and these are things that you know people would see potential opportunities to improve let's say inside of their performance analytics dashboards but then that next step of understanding the why behind some of those kpis that they were seeing wasn't always the easiest thing to get an understanding of process mining helps bridge that gap and get us to the why behind some of those kpis that we're seeing now how does this work well ever some of you probably already know this but every record on the platform creates an audit trail kind of like hansel and gretel leaving breadcrumbs and every time a record is updated it stores that change in the audit log and those are typically captured at the the task level these are turned on by default for something these autobots by default for something like 120 tables so you probably already have the data necessary to get started if you wanted to and then we have a mining engine that prepares the data at the relevant order data and handles all that heavy duty number crunching i mean just so you know that that mining engine that is essentially using the same machine learning infrastructure that things like predictive intelligence use so very often people start asking questions around performance impact the fact that we're using that machine learning infrastructure pretty much eliminates any performance impact on the main instance itself as we start doing that that's a question that typically comes up so i figured i'd call that out now we've taken that the mining process that we've built out here and we combine that with a growing library of common findings that guide customers to relevant insights through the process optimization analyst workbench there are a bunch of pre-built integrations with other platform capabilities like performance analytics benchmarks continual improvement and our predictive intelligence and that streamlines the kind of the the mining process and makes the insights to improvement process which is the important piece more efficient and complete some even like to say closed loop which is a necessary thing if we're looking to continually improve and not just make this a one-time thing so if we look at kind of optimizing processes on servicenow you can really break it down into four steps or four phases detect analyze improve and monitor right you're probably already using some of the platforms detection methods things like notifications on maybe sla breaches or performance analytics dashboards that have thresholds set to the metrics that you deem important to your organization or maybe you're even using the spotlight capability of performance analytics to help you identify opportunities or when things are going bump in the night that's the detect phase but once you detect something we move to the analyze phase and this is where process optimization optimization really dominates right with things like visualize process maps bottleneck root cause variation analysis as well as an integration with automation discovery which helps it uses machine learning to help identify automation opportunities for your organization so that's the analyze phase and we'll look a lot more dig into that more as we move forward in this presentation then once you find something right that's where the improvement process starts the platform is all about improvements uh whether we're automating decisions with predictive intelligence solutions or experiences with the virtual agent or manual steps with our in-platform rpa solution and sometimes maybe the improvement isn't an automation it's simply just setting up a coaching loop with the team member regardless there's always going to be plenty of opportunity out there to improve the processes in the organization we just need to identify where those opportunity exists then finally we want to make sure that whatever change we make is tracked right this is the monitor phase um and somebody's actually working on it very often you know i've found over the years people have found an opportunity to do better something cool inside of their data that tells them hey you want to do this they go tell somebody in a meeting of some sort or maybe they send an email and nothing ever happens with that opportunity and it just kind of goes wasted there this is where the integration with continual improvement management on the platform helps right because right from the analyst workbench you can create an improvement initiative and make sure that that gets assigned to somebody to follow up on and is tracked over time and then you use performance analytics to start tracking whether the change that you made is actually having the impact so detect analyze and prove monitor all four aspects in one platform closed loop as some would say so if we look at this how this these stages play out and let's say a throughput example that could play out across any number of workflows right we're always trying to increase the velocity inside of our organizations in terms of how things get done in the detect phase we can see maybe a significant increase in resolution times let's on a performance analytics dashboard or metric of some sort then in a single click from performance analytics which is not something that we're going to cover here today but right from the analytics hub you can drill into a process map for a subset of data a single click we could move over here to this analyze phase where you'll find the process map and we may identify that there's assignment group changes that are leading to a significant delay or there's some sort of ping pong going on from there we can decide to use predictive intelligence uh potentially to improve portions of the triage and assignment process and then we'd monitor our performance with a metric tracking maybe let's say first call resolution rate so all those things again one platform closed loop type of experience this throughput time is just one high level example across the workflows there are so many common findings and opportunities so many inefficiencies that exist inside of our organizations that process mining can help with as as part of that analyze step and what we've started to do is we've packaged these up into content packs essentially a library of pre-built finding templates to help you get started whether it be analyzing the performance of different intake channels or vendor performance comparisons bottlenecks causing sla breaches or process conformance issues but the the improvement opportunities are just endless once you start getting regular visibility into the reality of of your processes these are just the the top 26 that we found but we actually now have a list of up to 75 that we've identified through working with customers that we're in the process of packaging up and then here is what is currently in the itsm process optimization content pack right we're continually enhancing these on our own but i i can i can see a future right um as the servicenow community starts to get more and more familiar with process optimization i imagine there will be some who create and share their own sets of common finding rules and inefficiencies for for all of us to benefit from but this this itsm content pack has four pre-built continual improvement initiatives five process model definitions already built out and 13 finding definitions which are those things that kind of guide customers to areas they should be focusing on common inefficiencies that are out there i mean we'll look a little bit more at these finding definitions a little bit later on and actually on the schedule there's a deep dive session just for these finding definitions now if you look across all of the kind common findings and opportunities you can boil them down into three high-level buckets speed productivity risk removal so identifying opportunities to eliminate inefficiencies increase the speed in which things get done highlighting manual steps in a process that could in some cases be addressed by automation or enough there's steps in a process that simply could be removed to increase overall productivity and then providing visibility into process variability which needs to be controlled and better governed to remove risk for organization so what i thought i'd do now is i just do a quick little demonstration of the analyst workbench in action so you get a sense of what the end user experience looks like and then we'll transition into creating your first model so let's come over here and find my instance i'll have to log in again jay while i'm logging in here i don't know if you saw in the chat did anything come up in the chat from a question perspective yeah there is one question by the way what do you think uh so which database will be used for process mining instant database or something else like lots of database and can process mining be used to review slow job trends transcriptions scripts etc and what license is needed to install this so it's a bunch of questions there so how the process works and you'll see this in action right is um we extract the data from the audit table for whatever process that we're focused on and the subset of data that you'll you'll see when we configure it kind of how we can figure that the relevant data is pulled from the table and then sent off to the machine learning infrastructure to do the heavy number crunching and then the model is spit back into the instance and that's what we're using to drive the the visualization that you're going to see in a second here so that's that's kind of how that process works from a licensing perspective today the current situation is this is available as part of itsm enterprise and csm enterprise bundles um but we've we have identi identified that um that might not be this is something that most customers can get value out of if not all customers and we're looking at better ways of packing that up in the future uh so it's something to to kind of if you do have interest have a conversation with your your account executive about um and we can work to find a solution for you guys from getting it in your hands perspective and then in terms of the slow jobs and scripting it depends on kind of where those take place as part of the process it's not going to go all the way into the individual scripts themselves this is more at the the workflow and process level but if if those scripts are run as part of a process and you'll see when we get in here we'll be able to identify the steps in a process let's say from one activity to another one state to another that are the long-running states i mean as you drill down into those if there are scripts that kind of are run as part of that that's uh we might get some guidance there but i don't think it's going to go all the way into this get to the level of granularity that you might be asking about there all right now let's take a look here what i've come to is i've come to my process optimization workspace which says when you turn on the plugin this will be enabled for you this is just one way to get to process optimization going to the workspace itself and i'm going to come down here and i'm going to open up this project my incident process optimization project inside of here i've got a number of different tabs that give me information this first tab is the summary and insights tab and what this will give me visibility into some high level metrics that relate to the process that i'm interested in so in this case here we're talking about the incident process so i have key performance indicators around resolution times and first call resolution rate what happens to be a leading indicator for our resolution times then i've got my improvement opportunities and essentially what these are those these are those finding definitions or rules that i talked about a little bit earlier some of those pre-packaged inefficiencies that we commonly see um so things like hey multi-hop issues incidents going through one or two to two or more reassignments it's a very common thing what we can do when we do a mining is kind of bundle that um those records into it so in this case we've mined twenty two thousand records and six percent of them go through more one hops and that can just drive customers right into an opportunity to potentially improve or slow resolution time so incidents that took more than 12 hours between work and progress and resolved or unexpected transitions like incidents going directly to the closed state that's an area of opportunity or a conformance issue and for many organizations so we we pre-package a bunch of these for you per process um and then what we're doing is uh you have the ability to adjust these of course and then create your own uh findings um for the hypotheses that you have in your organization around inefficiencies that you might see we've got some variation analysis which will give us visibility into the various routes or routes that the work is taking so in this case here for the records that i mined i've got 3 200 different routes and then i've got my bottleneck analysis which focuses in on the individual transitions and the ones that are either high volume or long running so again some direction in terms of where i should get started at any point in time i can see one of these that is of interest to me and drill right into the workbench itself for just that subset of the data i'm moving quickly through this today because the next session will focus really on some of the analytical capabilities inside of here and how you can use the interface but i wanted to get a flavor today of what's possible before we create our first model now if i come over to the analyst workbench itself the first thing that jumps out is that map on the screen there of the process in this case we're looking at an incident process that's broken down by our state activities here i can get some basic model statistics so as i've mentioned already we're looking at about 22 000 records those 22 000 records have taken about 3 200 different routes the average duration is about three weeks and the standard deviation is three months which is a lot of variability which to me means there's a lot of opportunity to improve now you're saying to yourself maybe dan you said thirty 200 different roots it doesn't look like that there well that's because what we do by default is we isolate this down to the top 20 percent uh to kind of focus our attention on the areas that potentially have the biggest impact but if i wanted to blow this out and look at 75 of the paths it gets pretty messy pretty fast and i i promise you that anyone on your governance team is going to get pretty scared when they see this much variability inside of a process uh so typically we're going to stay at those those uh this 20 range but you have the ability to flux this and see the outliers the idea though at that 20 is this focuses on the high volume areas right instead of those one-off things that are important to understand but might not have the biggest impact if we're looking at say automation opportunities as i look at the map with each one of these transitions i get a volume metric here this is total occurrences right now but potentially i want to add a secondary metric and look at something like velocity of these things moving through each of the hops so now i'm seeing the ones that are going from i got 8 400 here um and they're taking on average about nine hours to move in this state here many people say to me dan this is this is new to a lot of people looking at a map like this or a process using a map like this and they're wondering where the opportunity or where should they start for me i always like to start on these sections where i see an arrow that's going back up the hill i'm back up here i always compare these maps for me they look like ski trails and i think back to when i was skiing quite a bit anytime i had to go back up the mountain that was typically a bad experience for me so i'm always initially looking for arrows that are going back up and that's usually an area of rework or an inefficiency that we might be able to focus on and i can come in here and i can click on any one of the nodes and i can start getting some details about that one specific transition in this case here i've got about 2400 records that move through that awaiting caller info to in progress node 18 of them are unique 1800 of them are unique that means to me like there's 500 that have done this taken this trip more than once which is a problem and one of them's even went back and forth at least six times or six times maximum and then i get some metrics around the duration of the things that are going through here and the total duration of 11 years that's crazy to me that's a lot of time or productivity that we can gather back potentially if we find a way to automate inside of here this is just one way to look at this another way that i can look at the the opportunities to use my bottleneck analysis and i can come in here and i can find that example in progress to awaiting caller info i can get the average duration standard deviation median duration but if i look at both of these in combination my in progress to waiting caller info and then back to in progress that total duration again it seems kind of crazy but that's 22 years of productivity that's being trapped in that back and forth where we're looking for additional information i want to figure out a way to kind of take some of that productivity back i can do things like isolating that one transition so i can come in here and i can say you know what i only want to focus in on things going through that awaiting caller info node so i can apply this and then i'll focus this down to just that and now what i've got is i've got about 7 100 so maybe a third of my cases are going to awaiting or tickets are going to weighted color info there and then i can start using things over here on the left these are my breakdowns i mean breakdowns i will cover this a little bit further and set these up but breakdowns are just different ways in which i can filter or analyze the records that i'm looking at right now so i can break this down by channel assignment group maybe i want to break it down by category or priority any structured data that you have you have the ability to start breaking these records down by inside of your map and if i start looking at these breakdowns let's say by channel which is a very common way to analyze people's data i can see that the largest amount of my records moving through this awaiting color info coming from self-service i mean their average duration is one week so you would think that we'd want to start there but for me what jumps out is this two-week average duration of tickets coming in via emails maybe i want to focus in on just those and you'd kind of expect that email's a very unstructured channel right so you would expect that we're not getting all the information we need right up front so we might have to go back and forth with the user to get that additional information so i can focus in on email and now if i want to just look at these that are coming through waiting for info i can do some additional analysis we've got in built-in root cause so i can see the leading influencers for just these tickets that are coming in via email that are going to waiting use the majority of them or 21 of them i should say are categorized as employee so maybe i want to focus on just those and then once i focus in on just those now i want to get down even further into the details and i can start using the platform's clustering capability to help me understand the unstructured data inside of those tickets that are coming in via email so i can start seeing like hey there's a large volume of these that have something to do with email and support and stuff like that let me just focus in and get down to the records here which you know if you have the rights to do you can in the platform and immediately can i can see a whole bunch of these have to do with somebody changing their email address as part of their profile which should be a pretty straightforward thing a thing for us to address uh with let's say a virtual agent conversation or potentially a service catalog offering maybe we might even have a service catalog offering already we just need to market a little bit better to kind of take advantage or drive some efficiency inside of the organization so just a real quick demonstration i wanted to get a feel for how you could potentially use from an end user perspective the analyst workbench before we went in and created our first um project and model together so you had some understanding of kind of some of the things we'll be setting up how they translate into the the end user experience all right swatch back over here to slide so um kind of what would we focus on here like these are all the different what i call building blocks of process optimization on the platform so today what we're going to focus on we're just going to focus on those ones at the top which are really the ones that we need when we're creating a model and then in future sessions we'll dive into some of the more analytical pieces there that you see on the list at the bottom now what makes process mining possible is event data an event data exists for each individual piece of work or record for example an incident case a request of some sort and as a piece of work moves through a given process it creates this event data which each step with each step it takes this event data at a minimum is going to have three values a record id so which piece of work is it that i'm focused on an activity so what's changed about that piece of work and a time stamp when did it change that event data is what is used to visualize what is actually happening within a given process now when it comes to servicenow the event data is stored in the audit table so turning on audit data collection is a configurable option for each table as i mentioned earlier it's uh turned on for about 120 tables by default and then if you have the rights to do so you can go into any record um and go up and you view the history so if you see there in the upper left you come to incidents you go down to view history i like to use the list the calendar view is a little bit hard to understand and you can get to the audit history uh for that record and kind of you can see that last visual at the bottom there that's the state transition for the record that we're looking at that's the data that we're using to create the maps that you're seeing on the screen here or that we just went through so that's that's the data piece what data we mine is all configured inside of a project projects are where we configure the source data and the filters for that data that we want to analyze and then besides configuring the source data we also configure a few other important pieces inside of our project like activities breakdowns and child or related tables that we want to map together so besides the source table an activity is the most important part of your project these are the different types of steps that can occur as part of your process think state or assignment group changes which are the most common but they can also be things like a knowledge article being attached or a task of some sort being created now you're required to have at least one activity configured for each project and you can have up to 10. but as you go through this process readability of the map is certainly something that you're going to want to consider i always guide folks to say you know what for your first time maybe just use state as as your activity just to get you comfortable before you start adding other activities there because readability is really important when you're when you're going through some of these maps as you saw they can get kind of crazy kind of fast now you can choose to group a given activity so this is useful for situations when you don't need to see each individual change for an activity but you want to know that a certain type of activity took place so for example you may be interested to know that assignment group change but you don't need to know the details of each individual reassignment or you may want to know when a knowledge article was attached but you don't need to see each individual knowledge article that was potentially attached so those are activities we'll configure those the next piece is your breakdowns right as you saw those breakdowns the actors while you were analyzing the map but they also give you statistical insights into the data because you saw some of those percentages and you can start sorting things by volume and velocity inside of there you're allowed to configure up to 10 breakdowns and these can either be choice or reference fields um so if you're not familiar overly familiar with the platform when we say choice and reference fields think like when you're viewing a form anything that's a drop down list or has a little magnifying glass next to it we don't allow people to create breakdowns on on string fields because that's uh just poor data it's just for data analytics practice i mean a question that typically comes up when we say you can have up to 10 breakdowns breakdowns are the high level grouping so channel assignment group service priority right so that's there inside of that um the number of elements like so that the number of different channels or the number of different assignment groups there's no real limit on those but you know you want to be mindful of kind of what you can consume as a human being and use for for analytical purposes and then specifically when using those breakdowns on reference fields you have the option to filter the breakdown so let's say you want to create a breakdown on location or assignment group or maybe even configuration item but you want to focus those breakdowns in a certain area you can add filter conditions to the breakdown values to make sure that they're a little bit more focused so those are the building blocks for the things that you need to understand for us to go and and create our first project so let's let's go create our first project and then once we do that we'll pause i see some things in the chat some questions that we might want to answer so we're going to come in here and we're going to create a new project and i think this is a super powerful view this ease of use of getting us from data to to a map is i think one of the biggest advantages you get by using in platform so we're going to call this one process optimization academy and then we can give it a short description this is for the first academy session you can configure goals that you align to these projects if you have organizational goals templates is something that we'll cover in the future but you might want to set this up as a template that gets applied when somebody mines from let's say a performance analytics metric but that's a future topic for a future session here you're going to select a table now you can choose a table or a different data source for those that are familiar of what we call report sources where you've got some pre-configured filters that you use over and over again you can set those up here as a data source and then i can come in here and i'm going to say i want to use the incident table because that's where i know i have my demo data and then typically best practice is you want to start mining closed records things that have completed their their process it's just the best practice you don't have to do that that's typically how we guide folks and if i hit preview here i'm going to see that i got about 30 000 instance incidents that match or that have been closed now another best practice and something that often comes up in composite conversation is how much data can i mine really there is no boundary um upper or lower that you would want to some people kind of think go back to the predictive intelligence guidance that we used to give by saying hey for predictive intelligence classification model you want to have a minimum of 30 000 records that doesn't exist here what you really need is to get an understanding of what would be a relevant set of data for you to mine for the given workflow or process that you're focused on is that a month's worth of data is it two months is it a week really it's what is relevant for you we don't care in terms of volume now i would say um because you know you want to be looking at processes as they exist in the current state usually what we guide people to is let's say let's just take the last month's worth of data so tickets that have closed in the last 30 days is a nice starting point and now this is demo data so when i hit preview i'm going to have zero records so i'm going to get rid of that that filter but that's kind of a best practice guidance the other thing that you can do here you can come and say you can start linking these to related tables so for example if i wanted to start filtering and just look at tasks or slas that have been breached i can come in here and i can say i want to use this related table to say actually i should do sla i think sla task and on this table i want the records that are so escalate task is attached to incident right and if i look for breached and we'll say has breached is true so now i want to start mining my incidents that have closed but also there they have an slas test that is breached so you have that ability to start working with related tables as well inside of here those of you familiar with the condition builder on the platform probably already knew that was possible once i set up the data that i want to use in this case i want to get back to my 29 000 i'm going to save this and then really we only have to do one more thing one more thing to get going and that's add our activity definition so i'm going to come in here and i'm going to say new and i'm going to add the most common one that you're going to add from an activity definition perspective is state and i'm going to say add an initial activity so it'll start with the when the record was created and make sure that we capture that data as well and we'll hit submit at this point we could be done we could start mining right now but i want to show you a couple of other things we'll add one more activity definition in this case let's add our assignment group and then from an assignment group perspective i might want to group those activities i could link an additional process here so if this was let's say a change was created i can come in here maybe i want to kind of link that processes from the process map you can look at the the pro it'll mine the the change path as well for the incidents linked from here we won't do that today we'll just hit submit and then let's create a couple of breakdowns some additional filters that will show up on the map so i'll come here to my breakdown definition hit new and we'll click on select and we'll add let's say category as a breakdown and we'll add one more remember channel is one of the more popular ones so let's add channel channels or contact types i might call it that's it we're good if i wanted to configure some additional tables to link to this i could do that here with child definitions if i wanted to configure who can see this process map i would do that here and then i've got some versioning and logging data that would show up as well and then what i'm going to do here is i'm going to click generate model sample and this is something best practice every time you're trying to model for the first time this is just going to take a subset of the 36 first 3 600 records or a subset of 3600 records to mine um in interest of time here today i think that's the best for us as well and this usually takes about three to five minutes so let me while this runs i'm just going to come in here i'm going to look at some of the questions that we've gotten we'll answer some of these question from phil are the the kpis used on the workspace defined just like any other kpi in the platform or they define special just for po great question so the kpis that you saw on that summary and insights page what those are um those are standard performance analytics key performance indicators and you would define them uh just like you would any other kpi on the platform or performance analytics metric on the platform by default we have inside of that content pack i was talking about set up that incident resolution one or resolution times and first call resolution rate as the metrics for in any time you're looking at the incident process but that is configurable you would use the ui builder ui builder interface to adjust those kpis or put new ones in there or add additional kpis that page is configurable via the ui builder um and then remember the safe harbor slide we're working to make that easier for you to configure moving forward and a little bit more flexible as well those are things that are on our roadmap for sure now the next question here one you'll see that my mining is already finished so that's a good thing so next question is about kind of setting thresholds or alerts to notify when something is out of bounds and really subjective based on your analysis so um so one of the things that's important to remember is there's those finding cards are the things that we would use to potentially um set uh notify people that things are out of bounds right so your your your rules in terms of things conforming are going to be different than other folks we've provided some of those things that are common out of the box but you're going to have you have the ability to create your own finding cards to help you guide your guide people to things that are out of bounds so that's that's one way to go about it and then there's a question here around setting up the um process optimization for custom tables as well so good question um and this is again safe harbor so today this is focused on itsm and csm um but there's no technical limitation for you to be able to use it against custom tables that you've created on the platform it's more of a licensing boundary today but that is something that we're working to address and is scheduled to be addressed in the tokyo release where we will then open this up for you to use process optimization against hr related tables as well as tables created via app engine or creator workflows again safe harbor our goal is to make process optimization available to all workflows that are running on the platform it's just we have to go through some some gates here at servicenow to kind of get uh make it available that way but that's that's our goal and as i said in tokyo the plan is to add hr and create a workflows um for to this capability so as you saw by the way we were answering questions this completed and i can immediately now go to the workbench for that data that we set up so that's that's what it takes like i said one of the things that i think is a huge advantage of taking an in-platform approach to using in-platform process mining is the ease of use from a data perspective right we know our data all we did there is configured a couple of fields in a form we didn't need to go to a data and analytics team or any of those types of things we just said this is the table that i want here's the activities that i want mine and within a matter of minutes i now have my process mapped and i can start doing all the things that you saw me do before in terms of analyzing the data it's that straightforward which is kind of why i wanted to start here today with instead of diving into all the different analytical capabilities inside the analyst workbench just showing you what it takes to get started because you have the right to turn this on in your subprime instance and mine up to 4 000 records today uh with uh just kind of as a trial or try it out so in terms of what do we do next now uh first thing i'll say is now that you know how to do it and how easy it is to set up and run in mind i would suggest next steps there is a process optimizations essentials course that's available it's self-paced on the now learning site so go to the now learning site takes about 60 to 90 minutes depending on how fast you read and click through some of the screens that are there but it's it's on demand no cost process optimization essentials it'll get you all the information that you need above and beyond will be covered here today then in your subprime environment you need to activate a couple of plugins process optimization is the main plugin and then we've got content packs for itsm csm telco and financial services for most of you it's probably going to be beneficial to turn on the itsm process optimization contact or the csm monetization content pack but you activate those two plugins and you can do what i just did in a matter of a few minutes and start seeing your data mapped in a way that you've most likely not seen it mapped before a couple of other resources that i think are important to call out for you we do many of you probably found this session via our process optimization community forum but if you go to community.servicenow there is a process optimization form that has a lot of information in terms of getting started i mentioned those common inefficiencies that we've seen they're all listed out up there there's a process optimization faq that's got a lot of answers uh on that page in terms and also links to other content so this recording we went through a lot of the stuff that we would typically go through in an initial conversation with somebody to help them understand what process mining on the platform is but we also have this recorded in in other places so there's an overview presentation and demo recorded out there if you wanted to share this this information with others inside of your organization so lots of other resources that are out there most of them you can get to from that process optimization community all right at this point um we went through a lot of stuff but let's open it up for some additional questions so if there's anything that you want to ask about what we covered here today please put it in the chat or if you feel comfortable coming off mute to ask go right ahead don't all jump at once hey this is lee can you hear me yes i can hey um so how soon would you do this after you installed ikea sam or csm for the very first time did you do it within 30 days i guess you could crank this thing up right yeah so that's that's a very common question and i think it's a very good one it's and the way that i always think about it is is how soon is too soon to know that you have an opportunity to do better um because we used to have this kind of this kind of question comes up a lot with performance analytics as well it's like well when do you start well with metrics one of the things is you want to start thinking about what metrics are important to you in parallel as you start to implement but with this is you know maybe two weeks maybe a month but as soon as you start using the system for real and you have actual users creating tickets of some sort maybe incense or cases you're going to want to start taking a look at how people how people are conforming to the process where are they veering off where do we see a bottleneck because the the earlier you can catch those things the earlier you can start making tweaks to whether it be your intake channel or whether it be how you're routing some of the cases it's to me it's you know it's it's i think weeks weeks is a is the probably the best way to answer that in weeks can he equal a month did that help lee yeah yeah i did thank you perfect anyone else got uh some questions any thoughts or any ideas for topics you want to cover in future sessions i'll just actually please keep asking questions but i will bring up um this screen here uh as you're thinking about your questions uh so today really what i i wanted to make sure that we got started with kind of having you see that ease of use and what it takes to create a model because that i think now that you all have this access to this video you can go home and or maybe right in the desk you're sitting in right now um and kind of turn on perhaps optimization and try to mine your data and then for the next session in two weeks i thought we'd spend more time on some of the different uh analytical capabilities in terms of the process map itself and navigating that bottleneck analysis variation analysis root cause some of those findings so all the different ways that you can interpret the data and find opportunities we spend time on that um and then as august rolls around uh we could talk i have a session on here in terms of what's coming in in the tokyo release so you get some visibility into kind of some of the improvements that they're making that we've making and there's some really powerful things that we're doing from an end-to-end workflow perspective and getting you visibility into processes like let's say hr onboarding that have several subtests uh that are often combined with that um and then we'll talk about i've got a session specifically on those um finding rules so as somebody was asking one of the questions in terms of hey can you set um control limits and thresholds and alerts and notifications most of that today is done excuse me as part of those finding rules but of course i'm interested to talk more about that topic with you to understand exactly what you're trying to to get to from an experience perspective because one of the things about these sessions is they're also uh we're i hope to get some input from you in terms of things that we might not have thought of from a product perspective that we can use to enhance it and then lastly uh something that we're hearing more and more is people interested in starting or spinning up uh establishing a process mining center of excellence so we've done that here at servicenow ourselves and other customers are starting to do that so we've got some thoughts that we want to share around how we've done it and how we're seeing others do it so those are kind of what we have from a roadmap perspective in terms of these sessions that will have every happen every other friday all right i hope i gave you some time to think of some additional questions anybody else got any other questions before we wrap up here today all right so last thing this uh academy for process optimization this uh you know we stole this concept from some other great academies that exist out there and if you didn't know that they existed um there's a an academy session that's just for the platform itself and capabilities in the platform a virtual agent academy the granddaddy of them all the the platform analytics academy as well as a mobile academy so there's there's other sessions just like this that are out there for you if you're interested in those other topics i mean like this this is the process optimization academy uh again coming back to the quote we started with do the best you can until you know better than when you know better do better hopefully now you know a little bit more and you'll take the opportunity to try this out um try this out on your own uh if you've got questions that we didn't answer here today use the the community forum i monitor that on a daily basis and i will answer your question as quickly as possible that's out there and that's where you would go to get some additional information and register for future sessions with that one last chance for questions before we we call it a day awesome well then i thank everyone for their time and look forward to seeing you back here on future process optimization academy sessions thanks dad yeah thank you dan thank you so much then take care